Constant Volume Combustion Chamber Research Articles

Understanding of combustion characteristics and NO generation process with pure ammonia in the pre-chamber jet-induced ignition system

As a favorable hydrogen carrier, ammonia (NH3) has great potential to achieve the goal of carbon neutrality in engines. However, it has several limits on the combustion in engines such as low laminar burning velocity (LBV), high ignition energy, and high nitrogen oxide (NOx) emission. It needs to employ an appropriate method for its utilization in internal combustion engines. The pre-chamber jet-induced ignition combustion mode which has been widely studied is an effective way to enhance the ignition and improve the combustion process of the mixture in the main chamber, especially under the lean-burn condition. To investigate the combustion effectiveness and NOx generation process of ammonia in this combustion mode, a three dimensions (3D) computational fluid dynamics (CFD) model coupled with a reduced ammonia reaction mechanism was established based on the constant volume combustion chamber (CVCC). The visualization experiments with jet-induced ignition combustion were conducted and used to validate the accuracy of the CFD model. The effect of equivalence ratio swept from 0.65 to 0.9 in the main-chamber was numerically studied in detail. The results indicated that the pre-chamber jet-induced ignition combustion mode can effectively improve the low flammability of ammonia and accelerate the combustion rate in the main chamber. The study of equivalence ratio sweeping demonstrated that a higher equivalence ratio leads to a higher combustion rate and lower NO and NO2 products, simultaneously. The increased equivalence ratio with more NH2 and NH radicals is beneficial for the NO reduction reaction and inhibits the oxidation reaction of NO → NO2. It reveals that a relatively high equivalence ratio in the lean-burn condition might be favorable for ammonia combustion.

Experimental and numerical study of the effects of N2 dilution and CH4 addition on the laminar burning characteristics of syngas

The study of combustion characteristics of complex biomass syngas under changing conditions is helpful for its practical utilization. In this paper, the effects of N2 dilution (dilution rates from 0% to 60%) and CH4 addition (mixing ratios from 0% to 20%) on the laminar flame properties of H2/CO/N2 syngas were researched. The laminar burning velocities (LBVs) were obtained by constant volume combustion chamber system. Five common mechanisms were used for numerical simulation using the CHEMKIN software package and the results were compared with the measurement data. The Li mechanism that was well matched with the experimental data was selected for kinetic analysis. The results indicate that the LBVs of H2/CO/N2/air mixtures decrease due to the dilution of N2. The addition of CH4 decreases the LBVs of H2/CO/N2/CH4/air mixtures and the LBVs of fuel-rich mixtures decrease faster. Sensitivity analysis illustrates that R29 is the most important branching reaction affecting LBV in H2/CO/N2/air premixed flame, the net reaction rate of all branched reactions and the mole fraction of H and OH in the mixture decrease due to the increase of N2 content. After adding CH4, the dominant branching reaction becomes R1. Due to the increase of CH4 addition ratio, the concentrations of H and O decrease, but the concentrations of OH and CH3 increase slightly. The chemical effect of adding CH4 increases LBV, but the physical effect decreases LBV. The reduction in the proportion of chemical effect is the reason for the faster reduction of LBV in the fuel-rich region. Finally, the formation of NO was analyzed in detail using the GRI 3.0 mechanism. The results show that N2 dilution reduces the mole fraction of NO, but CH4 addition increases the mole fraction of NO. Pathway analysis shows that N2 dilution and CH4 addition decrease and increase the effect of thermal mechanism on NO generation, respectively.

Methyl pentanoate laminar burning characteristics: Experimental and numerical analysis

This study investigated the laminar combustion characteristics of methyl pentanoate (MPe) using the constant volume combustion chamber at the initial pressures and temperatures of 1, 2, and 4 bar and 423, 453, and 483 K, respectively over the equivalence ratios ranging from 0.7 to 1.4. The experimental and simulated peak pressures showed a good trend irrespective of the disparities between both results. The experimental laminar burning velocities agreed well with the simulated laminar burning velocities. Also, laminar burning velocities at the initial temperatures and pressures up to 573 K and 8 bar, respectively were estimated using a deduced empirical formula. Last but not least, the main pathways for the decomposition of MPe were MPe → MPe2J → MPeMJ → MPe3J.

Experimental study on particle microstructure of diesel/biodiesel blend fuels measured by SAXS and HRTEM techniques

In order to investigate the mechanism on the microstructure of soot by adding biodiesel to diesel, the soot of biodiesel/diesel mixed fuel was collected by a constant volume combustion chamber, and the microstructure characteristics of soot were studied by SAXS and HRTEM techniques. The results showed that the added biodiesel reduced the delay of ignition and advanced the peak time of soot generation. The particulate matter in soot underwent a deep internal and external oxidation process. The accumulation of internal oxidation products forms a certain filling effect, which affects the subsequent oxidation rate. The microstructure of the single particulate matter was further refined to a cylindrical-like 3D sub-model filled with voids. The improved sub-model optimizes the DPF pressure drop model and provides a potential research basis for the DPF catalytic oxidation and regeneration model.

Effect of wall parameters on impinging combustion and soot emission characteristics of heavy-duty diesel engine at low temperature

The effect of wall parameters on combustion and soot emission characteristics remains to be revealed so far, especially in the case of spray impingement at low temperatures. Therefore, the visualization experiments are carried out in a constant volume combustion chamber using the Mie scattering and direct photography techniques, and the two-color method was used to extract the flame temperature and soot volume. The results show that increased wall distance and wall angle contribute to fast and stable ignition, but result in more soot production. In addition, the lower the ambient temperature, the more prominent the above characteristics are. At Tamb = 820 K, as the wall distance increases from 40 mm to 60 mm, the liquid spray impingement changes to the vapor impingement, so the ignition delay shortens, and the flame area and natural luminance increase significantly. At lower Tamb = 770 K, a complete misfire is observed at Lw = 40 mm but the ignition remains stable at Lw = 60 mm. The variation of the ignition characteristic parameters with the wall angle is similar to that of the wall distance. Under Pinj = 40 MPa, as the wall angle increases from 0° to 70°, the time to reach luminance saturation advances from 0.60 ms to 0.27 ms. Under higher Pinj = 100 MPa, a complete misfire occurs at θ =0° but bright flames are observed at θ =30–70°. With the increase of wall distance, the mean flame temperature increases due to reduced wall cooling. Coupled with the expanded flame area and combustion duration, the high-temperature region (>1800K) of soot distribution and KL factor increase significantly, especially in the range of Lw = 40–50 mm.

Experimental study of soot particles characteristics for hydrous ethanol diesel emulsified fuel under diesel-like conditions

The emulsified fuel of the commercial diesel fuel and 20 wt% water-containing ethanol with 30% volume fraction of hydrous ethanol was burned in a constant-volume combustion chamber (CVCC). At varying sampling locations along the axis of jet flame, soot particles are thermophoretically sampled and analyzed using high resolution transmission electron microscopy (HRTEM). Many small inception particles bulging from the surface of large “liquid-like” particles are observed at the upstream of jet flames. As these particles move toward the downstream, many growing immature primary particles gradually appear on those large “liquid-like” particles' surfaces, then carbonized particles fuse together and form larger soot aggregates, the large soot aggregates finally crack to generate many smaller particles. During this process, the diameters and the radius of gyration of soot particles decrease steadily, while the fractal dimensions have the opposite trend. The chaotic inner structure of primary particles progressively becomes an ordered structure. These results possibly indicate a new formation mechanism of soot particles; Compared to diesel fuel, soot particles derived from emulsified fuel are much immature with higher oxidation reactivity. The two-color images indicate the potential of reducing soot emissions by using hydrous ethanol diesel emulsified fuel increases the flexibility in controlling engine emissions.

A computational framework combining a basic turbulent combustion model, self-similar chemistry behavior, and a wall thermal law adapted for large-eddy simulations of a cyclically operating constant volume combustion chamber

The development of a computational tool capable of simulating and completing the experimental data obtained on a constant volume cyclic combustion (CV2) bench available at the Pprime Institute is the major objective of this study. The main difficulty is that, no matter how good the models used or the mesh refinement, specifying the boundary and initial conditions for the simulation of a specific cycle independently of the previous cycles requires an unattainable amount and accuracy of experimental data. In addition, high-fidelity simulation of all previous cycles is too costly. Developing a modeling framework compatible with coarse meshes for which the computational cost is reasonable is the first step necessary to simulate multiple cycles. However, the LES of CV2 presents numerous modeling difficulties due to the interactions of turbulent flow with fast and slow chemical reactions, compressible effects, walls and heat losses. Consequently, the wide spectrum of unresolved scales in the presence of so many correlated physical mechanisms leads to the introduction of several modeling parameters. Specific modeling developments are provided in this paper for turbulent combustion in the presence of residual burnt gases and the database takes into account the self-similar nature of the thermochemical properties. A detailed time-function for the inlet and outlet mass flow rates and a new blending function for the wall heat fluxes are also introduced. The sensitivity of the numerical results to the values of key modeling parameters such as injected air mass mair, ejected mass mout, segregation rate S∈[0.7;0.8], wall heat losses parameter κT∈[0.41;2.36] is studied. A quantitative comparison of the results and computational data is performed based on the pressure and heat flux signals in the case of propane-air combustion. The chamber pressure and temperature of air is 1 bar and 307K for the initial cycle. The calculated pressure profiles are in very good agreement with the experimental results. The values obtained for the calculated pressure peaks respectively for the first three reactive cycles are 17.4, 13.1 and 14.4 bars. For the experimental pressure peaks, the values obtained are 17.8, 14 and 14.6 bars which shows the validity of the modeling strategy. A qualitative comparison of the numerical results with other measurements, i.e PIV, flame visualization and MIE tomography, is also presented. The numerical results are then used to describe the flow during the successive phases, i.e injection, combustion and exhaust, of one specific cycle. Finally, the numerical pressure signal and wall heat flux obtained from the simulation of a sequence of ten cycles is compared to the experimental data. This last result demonstrates the ability of the LES tool to study the cycle-to-cycle variability observed in CV2 but further work is now required.

Numerical simulation of methane-hydrogen-air premixed combustion in turbulence

3D modeling is based on an experimental platform named turbulence constant volume combustion chamber (CVCC). The turbulence model applies the realizable k-ε model. The combustion model applies the premixed combustion model. Simulation of methane-hydrogen-air turbulent premixed combustion is realized using the model. The initial condition is T0 °C = 298 K, P0 = 1 bar and the gas equivalence ratio is φ = 1. The characteristics of flame are analyzed by modifying the turbulence intensity (uRMS = 0 m/s, 0.7 m/s, 1.3 m/s, 2.0 m/s)and hydrogen fraction (R (H2) = 0%, 10%, 30%, 50%). It is calculated that in the process of spherical flame combustion and propagation, the pressure increase rate increases first and then decreases. Turbulence intensity and hydrogen fraction have similar effect on the flame propagation process. If the turbulence intensity is used as an independent variable, the flame propagation rate increases with its increase. When the flame radius is 30 mm, the flame propagation velocity of 2.0 m/s turbulence intensity is about 2.7 times that of laminar flow. Due to which both the maximum combustion pressure and the maximum pressure rise rate are monotonically increasing. As it increases, the combustion pressure rises and falls the fastest. As the R (H2) increases, the flame propagation rate can significantly increase. When R (H2) is 50%, the maximum velocity is 4.82 m/s, which is about 1.5 times that of the maximum of pure methane. When R (H2) is high, the greater the maximum pressure is, the shorter the time for the maximum pressure to occur will be. When R (H2) is 50%, the maximum pressure is about 9 bar, which is 2.25 times of the force of pure methane combustion.

Experimental study on multi-channel ignition of propane-air by transient repetitive nanosecond surface dielectric barrier discharge

Multi-channel ignition by nanosecond surface dielectric barrier discharge (nSDBD) offers potential for efficiency boost in spark ignition engines. In this paper, transient repetitive nSDBD ignition is proposed using a simple pulse forming circuit to make multi-channel ignition more feasible, considering that nSDBD can be enhanced by high frequency pulses and that convenient discharge energy control can be achieved by adjusting pulse number. Ignition experiments for propane/air mixtures were carried out in a constant volume combustion chamber under stoichiometric conditions at ambient initial temperature and pressure. Discharge channel transformation and flame propagation were captured by a high-speed camera. For 20 kHz negative pulses, discharge energy rose from 4 mJ to 8 mJ as pulse number increases from 1 to 12, and filaments grew from 2 mm to 12 mm. Partially overlapping filaments formed a tree-like discharge pattern with divergent roots and a concentrated middle part. Multiple flame kernels were generated in the middle region of the filaments and then greatly promoted by subsequent discharges, resulting in a petaloid flame. Ignition probability has been tested and the minimum ignition energy of repetitive nSDBD is obtained. The influence of discharge energy on flame propagation speed and combustion characteristics were analyzed.

A parametric study of autoigniting hydrogen jets under compression-ignition engine conditions

This study examines the flame evolution of autoigniting H2 jets with high-speed schlieren and OH∗ chemiluminescence optical methods in a constant-volume combustion chamber over a wide range of simulated compression-ignition engine conditions. Parametric variations include the injector nozzle orifice diameter (0.31–0.83 mm), injection reservoir pressure (100–200 bar), ambient temperature (1000–1140 K), density (12.5–24 kg/m3) and O2 concentration (10–21 vol.%). The jet ignition delay was found to be highly sensitive to changes in ambient temperature while all other parameter variations resulted in minor ignition delay changes. Optical imaging reveals that in most cases, the reaction front of the H2 jet initiates from a localised kernel, before engulfing the entire jet volume downstream and recessing towards the nozzle. The flames attach to the nozzle, except at the lowest ambient oxygen condition of 10 vol.% O2 for which a lifted flame is observed. The H2 diffusion flame length shows a dependence on both the mass flow rate and the level of O2 entrainment that follows the same correlations as previously established for atmospheric H2 jet flames.

Experimental Investigation on Spray Evaporation and Dispersion Characteristics of Impinged Biodiesel-Butanol Blends

Abstract The objective of the investigation is to explore the spray evaporation and dispersion characteristics of impinged biodiesel-butanol blends at various n-butanol ratios (0, 10%, 30%, 50%) and ambient conditions. A total of 180 experimental cases were performed in a constant-volume combustion chamber. The liquid- and vapor-phase sprays were captured by backlight imaging technique and Schlieren imaging technique, respectively. Several macroscopic parameters were measured and discussed, including impinged spray structure, width, height, and area. Some novel parameters are derived to analyze spray evaporation and dispersion. Results show that biodiesel blended with 30% n-butanol transits better from liquid-phase to vapor-phase compared with other blends, displaying rapid liquid-phase evaporation an steady vapor-phase dispersion. After wall impingement, an increase in the ambient pressure or temperature hinders the liquid-phase dispersion in the vertical direction significantly, leading to a rapid decrease in the height of the impinged spray. The vapor-phase diffusion rate in the horizontal direction is about four times the rate in the vertical direction, and the rate ratio is slightly affected by ambient conditions and injection pressure. Compared with the free jet, the impinged spray is not beneficial for liquid-phase evaporation and vapor-phase dispersion, presenting larger liquid-phase area and smaller vapor-phase area. However, impinged biodiesel blended with 30% n-butanol displays better spray evaporation and dispersion.

Non-Reacting Spray Characteristics of Gasoline and Diesel With a Heavy-Duty Single-Hole Injector

Gasoline compression ignition (GCI) is a promising combustion technology that could help alleviate the projected demand for diesel in commercial transport while providing a pathway to achieve upcoming CO2 and criteria pollutant regulations for heavy-duty engines. However, relatively high (i.e., diesel-like) injection pressures are needed to enable GCI across the entire load range while maintaining soot emissions benefits and managing heat release rates. There have only been a limited number of previous studies investigating the spray characteristics of light distillates with high-pressure direct-injection hardware under charge gas conditions relevant to heavy-duty applications. The current work aims to address this issue while providing experimental data needed for calibrating spray models used in simulation-led design activities. The non-reacting spray characteristics of two gasoline-like fuels relevant to GCI were studied and compared to ultra-low-sulfur diesel (ULSD). These fuels shared similar physical properties and were thus differentiated based on their research octane number (RON). Although RON60 and RON92 had different reactivities, it was hypothesized that they would exhibit similar non-reacting spray characteristics due to their physical similarities. Experiments were conducted in an optically accessible, constant volume combustion chamber using a single-hole injector representing high-pressure, common-rail fuel systems. Shadowgraph and Mie-scattering techniques were employed to measure the spray dispersion angles and penetration lengths under both non-vaporizing and vaporizing conditions. Gasoline-like fuels exhibited similar or larger non-vaporizing dispersion angle compared to ULSD. All fuels followed a typical correlation based on air-to-fuel density ratio indicating that liquid density is the main governing fuel parameter. Injection pressure had a negligible effect on the dispersion angle. Gasoline-like fuels had slower non-vaporizing penetration rates compared to ULSD, primarily due to their larger dispersion angles. As evidenced by the collapse of data onto a non-dimensional penetration correlation over a wide range of test conditions, all fuels conformed to the expected physical theory governing non-vaporizing sprays. There was no significant trend in the vaporizing dispersion angle with respect to fuel type which remained relatively constant across the entire charge gas temperature range of 800–1200 K. There was also no discernable difference in vapor penetration among the fuels or across charge temperature. The liquid length of gasoline-like fuels was much shorter than ULSD and exhibited no dependence on charge temperature at a given charge gas pressure. This behavior was attributed to gasoline being limited by interphase transport as opposed to mixing or air entrainment rates during its evaporation process. RON92 had a larger non-vaporizing dispersion angle but similar penetration compared to RON60. Although this seems to violate the original similarity hypothesis for these fuels, the analysis was made difficult due to the use of different injector builds for the experiments. However, RON92 did show a slightly larger vapor dispersion angle than RON60 and ULSD. This observation was attributed to nuanced volatility differences between the gasoline-like fuels and indicates that vapor dispersion angle likely relies on a more complex correlation beyond that of only air-to-fuel density ratio. Finally, RON92 showed the same quantitative liquid length and insensitivity to charge gas temperature as RON60.

Constant volume combustion chamber (CVCC) investigations of aerospace F-24 and Jet-A in low-temperature heat release and negative temperature coefficient regions

A study was conducted to determine the characteristics of Low Temperature Heat Release, Derived Cetane Number, and the Vibrations, produced by the combustion of aerospace F24 using Jet-A as a baseline in a Constant Volume Combustion Chamber. The Ignition Delay and Combustion Delay for F-24 were found to be 4.04 ms and 5.71 ms with a DCN resulting at 43.77. Jet-A was found to have a larger DCN of 47 with an ignition delay and combustion delay of 3.35 ms and 5.14 ms. F24 has an extended region of cool flame and releases more of its energy during LTHR than Jet-A, however has a much shorter NTCR than that of Jet-A. From the NVH analysis it was found that there exists a strong correlation between the ID and CD phases and their respective NVH spectrum. It was found that the magnitude of the ringing for Jet-A was 3x higher than that of F24. Meanwhile, the magnitude of vibrations produced by Jet A was approximately 0.5 g greater than those produced by F24′s combustion which correlates well with the higher peak value of AHRR for Jet A and also with higher ringing produced by Jet A compared with F24.

Flame kernel development with radiofrequency oscillating plasma ignition

In this paper a radiofrequency oscillating plasma discharge is investigated under various initial pressures up to 5 atm in a constant-volume combustion chamber. The oscillating plasma is suppressed by elevated pressure, both in length and branch number. The ignition performance under elevated background pressure is investigated, and the results are compared with spark events with a similar ignition energy. Under ambient conditions, the oscillating plasma discharge generates multiple streamers that are much longer than a spark gap, resulting in a much bigger initial flame kernel. Under elevated background pressures, fewer streamers with much smaller sizes are observed, thus the advantage of an oscillating plasma discharge over a spark discharge is compromised. Prolonged duration of an oscillating plasma discharge consistently demonstrates a positive impact on flame propagation speed, but neither prolonged duration nor enhanced discharge current has a noticeable impact on flame kernel growth for the spark ignition cases. Both oscillating plasma and spark are used to treat non-combustible propane–air mixtures under background pressures from 1 to 5 atm.

Effects of High-Concentration CO2 on Ignition Delay Time of 70% n-Heptane/30% Toluene Mixtures

In order to research the high-concentration CO2 effects on ignition delay time (IDT) of diesel surrogate fuel (70% n-heptane/30% toluene), a carbon dioxide effect (CDE) model is established, which considers fuel and ambient gas concentration, density, and temperature influence on autoignition under CO2/O2 atmosphere. Firstly, a chemical model of n-heptane/toluene is established, and the coupling, reduction, and simulation processes are carried out in chemical kinetic software with the IDT as the target parameter. Secondly, a constant volume combustion chamber (CVCC) visualization platform is built by incorporating a high-speed camera system and different working conditions are set in the CO2 volume fraction range (40%-70%) at 3.0 MPa and 850 K for an autoignition experiment. Thirdly, experiment and simulation results are discussed in air, 60% CO2/40% O2, 50% CO2/50% O2, and 40% CO2/60% O2 atmospheres, including the IDT, CO2 effects, temperature sensitivity, and OH radical rate of production (ROP). The results show that the CDE model well predicts the 70% n-heptane/30% toluene IDT under the CO2/O2 atmosphere and the average error in 60% CO2/40% O2 atmosphere is 5.29%. Besides, when the CO2 volume fraction increases from 40% to 60%, the CO2 thermal effect plays a leading role in the IDT prolongation and the OH radical ROP peak of R4 (O+H2O⟶2OH) decreases by 180%.

Experimental and Numerical Study on the Effect of Hydrogen Addition on Laminar Burning Velocity of Ethanol–Air Mixtures

To understand the effect of hydrogen addition on the laminar burning velocity (LBV) of ethanol–air mixtures, experiments were conducted in a constant volume combustion chamber with the high-speed schlieren photography technique. The experiments were carried out under the equivalence ratios (ERs) of 0.7–1.4, an initial temperature of 400 K, an initial pressure of 0.1 MPa, and hydrogen fractions of 30% and 90% by volume, respectively. The effects of ER, initial temperature, initial pressure, and hydrogen fractions on the LBV were investigated. Moreover, adiabatic flame temperature (AFT), heat release rate (HRR), flow rate sensitivity analysis, and ROP (rate of production) analysis were also performed. Results showed that LBV increased with increasing hydrogen addition and temperature but decreased with increasing pressure. The hydrogen addition significantly increased the HRR of ethanol–hydrogen–air flames. The sensitivity analysis showed that R5 (O2 + H = O + OH) significantly influenced the LBV.

An Experimental and a Kinetic Modelling Study of Ethanol/Acetone/Ethyl Acetate Mixtures

With the world’s energy resources decreasing, ethanol/acetone/ethyl acetate mixed fuel has the potential as a fossil fuel alternative or oxygenated fuel additive. In this work, the burning characteristics of ethanol/acetone/ethyl acetate mixed fuels including 3 pure fuels, 9 binary fuels, and 7 ternary fuels were studied at a temperature of 358 K, the pressure of 1 bar, and the equivalence ratios of 0.7 to 1.4 in the constant volume combustion chamber (CVCC). The burning velocities of the ternary fuels were compared at ϕ = 0.8, 1.0, and 1.4. The results show that the laminar burning velocities of the mixed fuels are affected by the contents of ethanol, acetone, and ethyl acetate. The Markstein length, Markstein number, and burning flux were also analyzed in this paper. Furthermore, a detailed chemical mechanism comprising 506 species and 2809 reactions was reduced to a skeletal mechanism including 98 species and 642 reactions, using the directed relation graph with error propagation (DRGEP). The experimental and the simulated laminar burning velocities were compared. The results of laminar burning velocities show that the relative deviation of ETEAAC 112 is approximately 17.5%. The sensitivity coefficients, flame structure, and reaction paths of ethyl acetate were investigated with the skeletal and the detailed mechanisms. It is found that the key reaction path is retained in the skeletal mechanism.

Influences of the pre-chamber orifices on the combustion behavior in a constant volume chamber simulating pre-chamber type medium-speed gas engines

The study aims to clarify the influence of pre-chamber (PC) configurations on the combustion process in the main chamber (MC) of medium-speed spark-ignition gas engines equipped with an active PC. A constant volume combustion chamber was prepared to simulate the chamber configurations of the gas engines. A high-speed shadowgraph was applied to visualize the torch flame development and the combustion process in the MC. Experiments were done by changing the charged gas in the MC, the number, and the diameter of the PC orifices. Combustion was most accelerated when the PC orifice configuration was set appropriately so that the adjacent torch flames would combine with each other. It was also found that the unburned mixture in the PC, which ejected prior to the torch flame, supported the penetration of the torch flame.

NH3 combustion using three-layer stratified fuel injection for a large two-stroke marine engine: Experimental verification of the concept

A unique liquid NH3 spray combustion concept for a compression ignition engine, which applies a "three-layer stratified fuel injection comprising a supporting fuel / liquid NH3 / supporting fuel (i.e., NH3 stratified injection)” concept, in which the liquid NH3 is sandwiched by the supporting fuels, is proposed for a large two-stroke marine engine. The supporting fuel is a liquid fuel which can be ignited and burnt easily, such as marine gas oil, marine diesel oil, biodiesel oil and synthetic diesel oil produced from renewable energies. The three layers were injected sequentially from a single nozzle, and a three-layered spray was formed. To verify the combustion concept, a constant volume combustion chamber and a stratified fuel injection system were developed to simulate combustion in the cylinder of a large two-stroke engine. We then confirmed that the stratified fuel injection system can very nearly realize a two- or three-layer spray, by measuring the energy injection rate during one injection period. In the experiments, the heat release rates and combustion gas compositions for the NH3 stratified injection were evaluated. From the experimental results, and using the combustion concept, we found that the fuel is ignited and burns up the NH3 in the NH3 stratified injection; it also suppresses N2O production via the combustion of the supporting fuel in the third layer. Consequently, it was found that the combustion concept is effective for liquid NH3 spray diffusion combustion in large two-stroke marine engines.

Experimental Investigation on the Combustion Characteristics of NH3/H2/air by the Spark Ignition and Turbulent Jet Ignition

ABSTRACT In the current work, the combustion of NH3/H2/air blends is experimentally investigated in a constant-volume combustion chamber in terms of two different ignition modes: spark ignition (SI) mode and turbulent jet ignition (TJI) mode with ammonia and hydrogen fueled in the pre-chamber. The results show that for the SI mode, the burning rate and propagation of premixed NH3/air flames are significantly accelerated by a hydrogen addition of only 3% (v/v). A cellular structure is observed and becomes more intense with the increasing hydrogen volume fraction. For the TJI mode, the hot jet generation and turbulent flame development of ammonia inside the main chamber are captured optically for the first time. The combustion process inside the main chamber can be categorized into three phases for both the TJI-ammonia and TJI-hydrogen modes. The cold jet that mainly contains the unburned fuel and air is observed under the TJI-hydrogen mode. The hot jet produced by the pre-chamber can effectively promote the ignition initiation and hence flame propagation in the main chamber. An accelerated burn rate of ammonia is obtained, and the TJI-hydrogen mode makes a stronger promoting impact on the combustion of the main chamber. Significantly, compared with SI-hydrogen mode, the TJI-hydrogen mode can bring a considerable promoting effect to ammonia combustion with significantly lower hydrogen consumption.